-
Notifications
You must be signed in to change notification settings - Fork 22
Issues: pytorch-labs/attention-gym
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
How to manually check if one position or row has correct masking?
#66
opened Oct 28, 2024 by
Leo-T-Zang
How to reason about efficiency of different score/mask mod functions
#63
opened Oct 22, 2024 by
alex-hh
How to do KV Cache with FlexAttention and BlockMask by slicing?
#60
opened Oct 21, 2024 by
Leo-T-Zang
What is the best practice to save and load a BlockMask object?
#58
opened Oct 20, 2024 by
complexfilter
What is the expected gpu memory performance drop wrt flash attention with block masks?
#54
opened Oct 19, 2024 by
arilato
CUDA OOM Issue When Using Approx Tanh with softcapping score mod
#43
opened Sep 18, 2024 by
kebijuelun
[Feature request] End-to-end transformer example with flex attention
enhancement
New feature or request
#42
opened Sep 16, 2024 by
vladkvit
Dynamic shape compilation support for flex attention with block mask
#33
opened Aug 28, 2024 by
SamGalanakis
Support varied input sequence lengths with a fixed block mask
question
Further information is requested
#31
opened Aug 27, 2024 by
tilmto
It seems that New feature or request
good first issue
Good for newcomers
visualize_attention_scores
can only visualize either mask-mod-only or score-mod-only
enhancement
#29
opened Aug 23, 2024 by
XinDongol
[flex_attention] Softcap perf questions
question
Further information is requested
#22
opened Aug 16, 2024 by
meshtag
Previous Next
ProTip!
Type g p on any issue or pull request to go back to the pull request listing page.